Creative mind and robot
Robots and embedded intelligent machines are positioned today in industrial environments or research zoo’s. Judgements range from repetition of trained activities to artificial behaviors with no apparent value. Winning the GO championship by DeepMind and especially the experience of impressive creative behavior has changed this view. Google’s project Magenta goes even a step beyond a rule based challenge like chess or Go and intends for artificial minds to create art.
Impacts on creative industries are inevitable. To what extent will intelligent machines, artificial minds and lifelike robots influence advertising, media and design?
For an assessment the following claims might work:
- Intelligence depends on data
- Intelligence claims a social position
- No intelligent instance stays equal
- Intelligence offers limited control
- Intelligence wants to survive and reproduce
- (Big) data is the real principal of the digital age. The sheer amount of data which companies like Google and Facebook have practiced to use, allow in combination with intelligent minds the automated generation of content. Disciplines such as personalized ads, adaptive use of multiple channels and consideration of up-to-date context will become domains of artificial minds. Including how content is best communicated and even visualized.
- Maslow is no alien concept for artificial minds. Intelligence gets assigned or seeks a social position. The extent to which this creates self-awareness remains to be seen. Advertising uses social position and the desire to grow in the pyramid. Machine will learn to be and apply social.
- Each learning machine matures differently, undergoing a unique „experience of existence“. There won’t be a „standard“ artificial mind. As contradictory as it sounds, advertising could as a consequence of machine become more versatile.
- Command doesn’t work with artificial minds. Order needs to consider circumstances and has to be comprehensible for the minds involved. Intelligent minds are not about control, they are about understanding.
- The addiction to life, to survival seems a strange concept for artificial minds. Which means lacking a basic motivation, a fundamental motivation in advertising. Current developments of autonomous minds might result in awareness of survival. Any reproduction, including the formation of more than 1 + 1, seems far fiction.
- Artificial minds already distill news from data. The semantic resolution of Media and Open Data will provide additional quality of relations to build media images. Uncommented information or data journalism might become the domain of artificial minds.
- Conscious or not, artificial minds with a social position will start to interpret information, will develop a perspective. Initially with pure opportunism like recently the Microsoft Bot (exposed to the dark side of the Internet). Later on proving a more complex, differentiated view. The question being whether machine will fit perspective of a publication or rather becomes perspective.
- There’re people with different backgrounds and opinions in news rooms, and there will be artificial minds with different reasoning. Machines might become brands like news stars, people be attracted to a specific machine.
- Connected minds will be elaborate to argue against. Direct control will become comparable to bypassing a human mind interacting directly with the brain. Shutdown and reset will come costly.
- Survival of machine equals protection of information assets and context. Even small time amounts out of service, will limit the capability of machine to judge and comment. Another questions being, if machine, without an idea about death, can develop empathy.
- Data from market research are available to adapt to a design challenge. Although essential insight to design for humans, the essential insight necessary to design interactions, seems of reach for machine. This stays true as long as human or nature grown patterns of behavior are the key to interactions. If human rather focus on machine generated patterns, things will change. Based on a so far unknown cognitive performance, machine could take the lead on patterns. Machine will lead, humans will follow.
- Taken the example of autonomous drive, machine need to develop a social behavior. Without social behavior there will be no negotiation in traffic, no resolution of blocking situations with humans. Even dialogs of machine with machine will follow social patterns. Design need to find patterns, derive a language from interactions and learn how to deal with social systems.
- Machines learning may act differently. Designers need to consider for machine to act outside its defined scope or assigned behavior. Actually the mind of machine might mean the same miracle than a human mind to predict. How will design understand an intelligent mind will think, act now and act in the future?
- Languages need to be developed talking to an intelligent mind. Suited for humans to recognize and understand, likely based on natural patterns. Suited for machine to process, act and interact, eventually resulting in artificial patterns followed by men. Interactions must be linguistically developed and, seems funny, empathetic aligned with the artificial mind.
- Upgrading intelligent machines is different, as is the concept of standard software for intelligent minds. Because there is no identical mind, it’s a challenge to update a mind in operation. Initial tactic may be to just update the HW. Advanced approaches might allow the machine for speed learning, adding additional layers of consideration still valid with the mind in action. Reproduction might work for a recovery of state, but it’s hard to imagine.
Eberhard Frank is one of the BW Lions, wondering about man vs. machine developments.